Repository navigation
Align Moli automation profile and explicit viewport preconditions - #28
Merged
lanyue-llk merged 2 commits intoSep 22, 2026
Merged
Conversation
This was referenced Sep 22, 2026
lanyue-llk
added a commit
that referenced
this pull request
Oct 10, 2026
* Fix Moli cohort harness infra and freeze version comparison * Summarize task outcomes in controlled Moli comparison * Exclude Chrome-ineligible diagnostics from Moli cohort * Make Moli layout policy explicit for automation cohorts * Route Schema probes to page target * Add task-scoped Moli layout policy with frozen assignments * Default unclassified Moli task contracts to real layout * Add three-state task layout requirements and automatic qualification * Reuse historical layout-off passes across engine upgrades * Replace auto layout with opt-in failed-case reruns * Refactor layout recovery and clarify remaining failure tracking * Version Moli failure cohorts and exported bad-case lists * Keep versioned input cohorts separate from run results * Keep all historical Moli failures independent of Chrome * Correct Chrome baseline probe conditions * Add verified try-layout cohort summary * Bind candidate cohort to Moli commit * Treat candidate commit as binary identity * Record candidate host provenance * Run Python adapters with harness interpreter * Measure process tree resources on macOS * Certify macOS Moli resource cohorts * Bind macOS resource sampling metadata * Account for try-layout retry resources * Report frozen Moli resource comparison population * Measure fixed Moli layout resource configurations * Summarize fixed layout outcomes and two resource runs * Allow official layout-on resource summary without an off run * Bind fixed-layout attempt evidence to source runs * Remove layout retry and retain only fixed layout runs * Fix cohort run identity and resource binary provenance * Scroll adapter targets into view before pointer operations * Align Moli automation profile and explicit viewport preconditions (#28) * Apply automation profile to Moli and set explicit viewport * test: cover automation profiles for supported engines * Bind cohort comparisons to physical result identities * Add visual overview of Moli benchmark infrastructure changes * Use module imports and remove obsolete Moli launch plumbing * Consolidate fixed layout documentation in running guides * Keep direct tool scripts and install their imports normally * Isolate cohort engine selection and reject foreign runner imports --------- Co-authored-by: kevin <xxx>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The browser-automation launch profile already enables --enable-automation for Chromium, but does not enable it for Moli. The viewport probe also assumes a platform-specific default width of 780 CSS pixels.
Add Moli to the existing task-scoped browser_automation profile. Set an explicit 780×600 viewport with DPR 1, preserve all three width assertions and add cssVisualViewport.clientWidth. No global launch switch, scoring exception or engine-specific expected value is needed.
This PR targets the existing harness PR #25 branch, where task-scoped automation profiles are defined; it does not duplicate that work. The viewport task is byte-identical to the frozen 12-case verification contract. Validation on this exact PR commit (82813ce): only the original 12 cases, three attempts per engine; Moli 36/36 and Chromium 36/36, run pr28_original_12. No other benchmark cases ran.
Moli implementation: lexmount/moli#731
Report, task definitions and evidence: https://artifact-site.lexmount.cn/v/WiF5KVJ_Ta2lgsaj3r81hCejInBl4u1f
Review illustrations and evidence
完整验证报告：逐项根因、Chromium 对照与原始证据
以下为 ImageGen 生成的机制示意图；执行结果与原始证据见报告。
评测前提：自动化启动配置与显式视口
启动参数决定自动化接口是否可用；视口断言必须建立相同的 CSS 像素与 DPR 前提。